Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
ICT Investment and Economic Growth in India: An Industry Perspective
Abdul Azeez Erumban
(The Conference Board, Belgium)
Deb Kusum Das
(University of Delhi, India)
Paper prepared for the 35th IARIW General Conference
Copenhagen, Denmark, August 20-25, 2018
PS16: Other
WEDNESDAY, AUGUST 22, LATE AFTERNOON
1
ICT investment and economic growth in India: An industry
perspective
Abdul A Erumban1, Deb Kusum Das2
Abstract
The role of information and communication technologies (ICT) in driving economic growth has been well established
in the literature. By reducing communication and transaction costs, and improving the quality of capital, ICT helps
firms improve their productivity and growth. Given her linguistic and engineering skills, India has been pioneering
in ICT exports, in particular export of software services since the 1990s. However, there is hardly any attempt to
understand how Indian industries have been taking advantage of the massive growth potential of ICT use in their
production process, looking into the experiences of different industries. This has been primarily constrained by lack
of adequate, disaggregated data on the ICT use by industries. While there are a few studies trying to understand the
contribution of ICT to aggregate economic growth, almost no study has attempted to unearth the role of ICT at detailed
industry level. This paper is a first attempt to construct ICT investment series for the registered or organized segment
of manufacturing industries in India, and one of the first few attempts that have made so far to build such ICT series
for the aggregate Indian economy. The study extends the capital asset database in India KLEMS to include ICT
investment, i.e. investment in hardware, software and communication equipment, in respect of different manufacturing
industries. The paper also provides preliminary estimates of the contribution of ICT capital to growth in aggregate
economy and registered manufacturing sector.
Keywords: India; economic growth; information technology; ICT; organized manufacturing; industry-
wise investment, aggregate economy
1. The Conference Board and University of Groningen, [email protected]; 2. Department of
Economics, Ramjas College, University of Delhi, India, [email protected];
Acknowledgements: This paper draws heavily on the India KLEMS project funded by the Reserve Bank of India.
The authors are thankful to the Reserve Bank, and many researchers who have contributed to the development of this
data. In particular, we thank K.L. Krishna for his comments and contributions. Comments by participants of the India
KLEMS Conference at the Delhi School of Economics in December 2016, in particular Ashok Jain, T. Rajeshwari,
Rajiv Mehta and S. Sahoo were quite beneficial to this work. Research assistance by Gunajit Kalita and Sreerupa
Sengupta are also acknowledged. Views expressed in this paper are those of the authors and do not reflect their
respective institutions. The usual disclaimers apply.
PRELIMINARY VERSION, not to be quoted without permission
2
1. Introduction
The role of information and communication technologies (ICT) in driving economic growth has
been well established in the literature (see for instance Jorgenson et al, 2011; Inklaar et al., 2008;
Jorgenson and Vu, 2005; Jorgenson et al, 2005; van Ark et al, 2003; Jorgenson, 2001).1 By
reducing communication and transaction costs, and improving the quality of capital, ICT helps
firms improve their productivity and growth. In Figure 1, we depict the ratio of ICT investment to
GDP in 81 countries over the period 1990-2015 against labor productivity growth. It is quite
evident that there is a strong correlation between the two. Even more important is that this
correlation is stronger in the emerging market economies, and remains strong even if we exclude
outlier countries. It is also evident that the emerging market economies are quite behind in ICT
adoption, compared to most advanced economies. This clearly suggests that these economies still
have significant potential for improving their productivity growth by investing in ICT.
1 For a while, productivity statistics failed to capture the effect of ICT, as the rapid declines in ICT prices and
information on ICT investment were not fully captured in official data. In 1987 Robert Solow even remarked, “[Y]ou
can see the computer age everywhere except in the productivity statistics”, an observation which is often considered
as ‘IT productivity paradox”. However, later studies which accounted for rapid declines in ICT prices using quality-
adjusted hedonic prices observed significant contributions from ICT (see e.g. Jorgenson and Vu, 2005). The debate
on price is, however, far from settled. For instance, a recent study by Byrne and Corrado (2016), argues that the official
US hedonic prices fail to capture the magnitude of price declines in ICT, and their alternative measures suggest even
larger declines.
3
Figure 1: Correlation between ICT investment and labor productivity growth, 1990-2015
Note: The red dots are emerging markets and the blue dots are advanced economies. The gray solid line is
the linear trend line for the entire 81 countries, and the red solid line is the linear trend line for emerging
market economies only. The red dotted line is the trend line for emerging market economies only, but
excluding outlier countries.
Source: The Conference Board Total Economy Database, November 2016
Given her linguistic and engineering skills, India has been pioneering in ICT exports, in
particular export of software services since the 1990s. Taking its early bird advantage, India has
maintained a relatively higher market presence in the software segment of IT industry (Saith,
Vijayabaskar, 2005). However, there is hardly any attempt to understand how Indian industries
have been taking advantage of the growth potential of ICT use in their production process. While
there are a few studies trying to understand the contribution of ICT to aggregate economic growth2,
the evidence on the ICT’s impact on growth at detailed industry level is even more limited.
Available micro level evidence based on plant or firm level data is suggestive of a positive impact
of ICT on growth and productivity in India’s organized manufacturing sector. For instance, Sharma
and Singh (2012), using ICT investment data from ASI, observe that higher ICT stock is associated
with higher levels of value added. Similar conclusions have been made by Joseph and Abraham
(2007), who observe a positive effect of ICT investment intensity on both labor productivity and
2 See Erumban and Das (2016) for a most recent analysis of ICT’s contribution to aggregate economy growth.
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
0.00 0.01 0.02 0.03 0.04 0.05 0.06
Lab
or
Pro
du
ctiv
ity
Gro
wth
ICT investment / GDP
4
TFP, even though the ICT intensity in the sector is low. Comparing firms in India and Brazil,
Commander et al (2011) suggest the importance of better infrastructure and labor market policy
for higher rates of ICT adoption. These policies might be imperative for increasing the ICT
intensity in India’s manufacturing, which is arguably low, as is evidenced by previous studies. Kite
(2012 and 2013) observe positive ICT effects on productivity, and in particular the use of
outsourced ICT makes larger contribution than in-house ICT.3 A recent paper by Navyashree and
Bhat (2016) analyzes the ICT investments and its growth effects in small and medium firms in
food processing industry in India. Using Prowess data, they demonstrate a positive impact of ICT
use on growth of small and medium enterprises. At a more aggregate level, Erumban and Das
(2016) provide estimates of ICT contribution to growth for the aggregate manufacturing. They use
ASI and NSSO data on ICT investment, respectively for registered and unregistered segments of
the manufacturing sector. Their findings suggest a lower contribution of ICT to growth in the
manufacturing, compared to the aggregate economy, thus implicitly suggesting a larger ICT
contribution in the services sector.
Two recent studies looked at ICT in Indian economy from a different perspective, which
are Mitra et al (2016) and Vashisht (2017). Mitra et al (2016) looks at the importance of ICT
infrastructure for productivity growth in Indian manufacturing. While their study does not have
any explicit measure of ICT capital (they have a measure of total capital stock based on CMIE’s
Prowess data, which is obtained using a perpetual inventory method assuming 7 percent
depreciation rate), they evaluate the impact of ICT infrastructure on productivity, with the ICT
infrastructure data being obtained from World Bank’s World Development Indicators. Vashisht
(2017) looks at the job impact of technology, where ICT is considered as one of the major
technologies. They find no evidence of technology destroying jobs, rather their evidence is
suggestive of it shifting jobs from intermediary skill to high skill jobs. While technology has led
to an increase in the capital intensity, it has not reduced aggregate employment in Indian
manufacturing. They use ICT investment data reported in ASI – the same data as we use in this
paper (see the discussion below) – in constructing ICT capital stock.
3 IT outsourcing in these studies is measured using data on expenditure on software and other professional services
obtained from PROWESS.
5
Thus the evidence on ICT’s impact on economic growth in India is quite limited. The major
reason for this, despite India being one of the major ICT service providers, is the lack of adequate
consistent data on the ICT use by industries. Since the onset of India KLEMS database, which
provides detailed asset wise information on capital investment at industry level, along with other
indicators required to analyze sources of growth, it is now feasible to examine the sources of
economic growth in India at detailed industry level.4 This study intends to extend the India KLEMS
data to include ICT investment, i.e. investment in hardware, software and communication
equipment, in the aggregate economy and in the organized manufacturing industries. While the
India KLEMS is an exhaustive database, covering all segments of the economy – be it organized
or unorganized – due to data limitations, the current version of the paper will analyze the role of
ICT in the organized manufacturing industries only, using detailed asset data from Annual Survey
of Industries (ASI). Note that the India KLEMS data is fully benchmarked to National Accounts
Statistics (NAS) data. The basic estimates of industry wise ICT investment data in this paper is
consistent with ASI, and is subsequently benchmarked to NAS data to ensure consistency. Future
extension may attempt to extend it to other India KLEMS industries, including unorganized
manufacturing and services.
Following the economic liberalization and large scale globalization policies in the early
1990s, India has benefitted from the second unbundling of globalization – the ICT revolution that
has radically lowered transmission costs (see for instance Baldwin 2016). As Erumban and Das
(2016) noted, Indian’s English language ability, engineering and programming skills accompanied
by lower labor cost helped the country catch-up quite fast, particularly in the software sector.
Existing evidence suggests that the ICT revolution has helped several Indian firms introduce
innovative entrepreneurship and governance model within the software sector (Arora and Athreye,
2002). Going beyond the software producing sector, the use of ICT is expected to reduce
communication and co-ordination costs in ICT using industries, and thereby help them improve
their productivity. Following previous literature, Erumban and Das (2016) summarizes the
channels through which ICT can contribute to economic growth, which includes: a production
channel, in which ICT producing sectors benefit from rapid technological change in the sector; an
investment channel, in which firms investing in ICT could enhance the contribution of capital to
4 See Das et al (2016) for a recent study that exploit this database to study sectoral productivity dynamics in Indian
economy.
6
growth, and a productivity channel, in which firms using ICT could improve their productivity.5To
understand these channels better, we need to have industry-wise data on investment in ICT, along
with output, employment and investment in other types of capital assets. While the India KLEMS
provides all but the ICT, this paper is an attempt to fill this gap.
The remaining of the paper is presented in four sections. The second section discusses the
methodology to estimate ICT investment for Indian economy. First it discusses the approach to
estimate aggregate economy ICT investment, and the trends in estimated ICT investment series.
Subsequently it presents the approach used to derive a series on investment in ICT for the
organized (registered) manufacturing industries covered by the Annual Survey of Industries (ASI).
The results are also discussed in the section. In Section 3, we discuss some practical aspects or
issues that need to be tackled while incorporating the estimated series of ICT investment into India
KLEMS growth accounting database. In section 4 we provide some preliminary results on our
estimation of the contribution of ICT capital to growth in aggregate economy and registered
manufacturing. In the latter case, following the India KLEMS practice, to keep full consistency
with National Accounts (NAS) data, the obtained ICT series for ASI sector has been normalized
to NAS aggregate GFCF data. In section 5, we provide a summary of main findings, and in the
last section we discuss the way forward.
2. Estimating ICT investment in Indian Economy6
There is hardly any official data on complete ICT investment series, covering hardware, software
and communication equipment, in India. Therefore, it is essential to estimate a series of ICT
investment compiling various sources of information. We follow a three-step approach in
estimating ICT investment for the Indian economy. First, we estimate the investment in hardware,
software and communication equipment in the aggregate economy, which is consistent with
available data from the National Accounts Statistics (NAS). Then we estimate industry-wise
investment for the organized manufacturing industries – i.e. 13 India KLEMS manufacturing
industries. In the third step, we combine the two series, by which the manufacturing sector
5 Also see van Ark et al (2011) 6 The discussion in this section on aggregate economy ICT investment heavily draws upon Erumban and Das (2016).
This paper is an extension of Erumban and Das (2016) in that it constructs ICT capital in detailed manufacturing
sectors, though confined only to the registered segment.
7
estimates are benchmarked to National Accounts data. In what follows we discuss the approaches
we follow to estimate aggregate economy series and organized manufacturing sector series
separately.
2.1 Approach to estimating aggregate economy ICT investment
The available information on ICT investment in India include software investment from NAS
since 1999-00, ASI’s ICT investment series in organized manufacturing sectors since 1998, NSSO
62nd round data on ICT investment in unorganized manufacturing, CMIE’s PROWESS firm level
data on gross fixed assets in hardware, software and communication equipment (1989 onwards)
and World Information Technology and Services Alliance (WITSA)7 ’s estimates on ICT spending
by broad sectors of the economy since 2000. In this paper, we make use of NAS, ASI and WITSA
data, along with investment flows by commodities reported in Input-Output Tables (to be
discussed later).
Software investment: National Accounts Statistics provides software investment by different
undertakings covering administrative departments, autonomous bodies, cooperatives,
departmental enterprises, household sector, non-departmental enterprises, private corporate sector
and public administration, since 1999. We have obtained this data disaggregated by industries from
Central Statistics Office (CSO), and this has been the basic series we consider as the benchmark
series of software investment for India. A comparison of CSO’s software series with that of ICT
spending data from WITSA suggests a substantial downward bias in WITSA software spending
data. For instance, for 2008, WITSA software spending to GDP ratio is 0.12 percent, excluding
consumer spending and 0.17 percent including consumer spending, whereas NAS software
investment/GDP ratio from National Accounts is 1.2 percent.
7 WITSA provides ICT spending data in a cross section of countries through their Digital Planet Report. See
http://www.witsa.org/
8
Figure 2: Ratio of investment in software to hardware (current price), aggregate economy –
the three approaches
Note: The blue solid line is the ratio of software/hardware spending in non-consumer segments, obtained
from WITSA for post 2000 period. This data has been applied to NAS software investment series to estimate
hardware investment for the post 2000 period. The U.S ratio (the purple solid line) is plotted on the right
hand scale. The blue dotted line assumes a constant software/hardware ratio (based on WITSA) for all
years, the red dotted line is a linear trend based on available data and the green dotted line is the trend in
the U.S ratio applied to 1999 software/hardware ratio in Indian economy.
Source: Authors’ computations, using data from WITSA, EU KLEMS, NAS, and Input-Output Tables.
For years before 1999, we extrapolate software series by applying software/hardware ratio
to measured series of hardware investment (measurement of hardware series is discussed below).
Here, we use three alternative approaches to derive time series of software/hardware ratio. The
first is to keep software/hardware ratio from WITSA in 1999 as constant; the second is to use a
linearly forecasted software/hardware ratio from WITSA; and the last and preferred series is to
generate a series of software/hardware ratio using the trend in software/hardware ratio in the
United States (see Figure 2 for the software/hardware ratio using these three approaches). The U.S
data has been obtained from EU KLEMS database.8 While this is still a rough assumption, it is
better than assuming a constant or linearly growing software/hardware ratio. The U.S, being the
8 We use the 2011 version of the EU KLEMS data, see www.euklems.net.
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.0
0.2
0.4
0.6
0.8
1.0
1.21
99
0
19
91
19
92
19
93
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
02
20
03
20
04
20
05
20
06
20
07
20
08
20
09
20
10
20
11
20
12
Constant Linear forecast U.S trend WITSA US ratio (RHS)
9
ICT leader, may provide a realistic picture of the required hardware/software ratio, particularly in
the early years. Hence we use the third approach – the trend in U.S software/hardware ratio.9
Hardware investment: National Accounts, however, does not provide investment in hardware and
communication separately from total machinery investment. In constructing a series of software
investment, once hardware data is available, de Vries et al (2010) suggest using the elasticities of
hardware to software investment, estimated using a fixed effect panel regression of software on
hardware and a set of control variables. We follow a similar approach to derive hardware
investment, but not using econometric techniques. We use software/hardware ratio from WITSA
ICT spending data (see the blue solid line in Figure 2). While doing this, we exclude consumer
spending. This ratio is applied to software series obtained from National Accounts, thus providing
us a hardware series for 1999-2011.
For years before 1999, hardware series is obtained using the trends in the hardware
investment series arrived at using a commodity flow approach (CFM)10. That is the benchmark
series for 1999 is extrapolated using the annual changes in the investment series obtained using
the CFM. In the CFM approach, total economy investment in hardware and communication
equipment can be estimated using the information on the total domestic availability of these goods
and its investment component. This requires the use of input-output tables, in combination with
NAS and trade statistics. We define the investment in ICT asset i as:
( )( )tititiIO
si
IO
si
IO
si
IO
si
ti EXIMYEXIMY
GFCFGFCF ,,,
,,,
,
, −+−+
= (1)
where GFCFi,t is the current gross fixed capital formation, Y is gross domestic output, IM is
imports and EX is exports – all for aggregate economy. Superscript IO refers to input-output tables,
9 Note that the software investment data – be it directly from the NAS, or obtained using hardware/software ratio
from WITSA or from the United States - does not capture pirated software used by companies, if any. While the use
of such pirated software would indeed contribute to firm’s output growth, it will never be reported by firms, and hence
is hard to capture. 10 See de Vries et al (2010) and Timmer and van Ark (2005) for a good description of the commodity flow approach.
10
i.e. for instance, IO
siI , indicates investment in asset type i (since we consider computer hardware
and communication equipment, i=1,2, i.e. hardware and communication equipment) in years
(where s is the benchmark year for IO table) obtained from input-output table. All other variables
without the superscript IO are time-series data obtained from the NAS. Following the previous
studies, we define industry 30 according to ISIC 3.1 (office equipment and machinery) as computer
hardware and industry 32 (radio, TV and communication equipment) as communication
equipment. There is no strict concordance between ISIC 3.1 and India’s input-output table
classification, and therefore, we consider the Indian IO sector office computing and accounting
machinery as hardware, communication equipment and electronic equipment including TV as
communication equipment (see Table 1). We obtain investment in hardware and communication
equipment, along with total domestic output, imports and exports for 6 benchmark years, 1983,
1989, 1993, 1998, 2003 and 2007 from input-output tables published by the CSO. This information
is used to compute the first part of equation (1). Then, using time-series data on gross output
obtained from National Accounts and exports and imports obtained from UN-Comtrade statistics
(see Table 2 for the concordance between Comtrade and ICT assets), we construct a series of
hardware investment using equation (1).For years before 1983 (i.e. the first bench mark I-O table),
we linearly forecast the domestic availability ratios (see the first part of equation 1), till 1970, so
that we can derive a long series of ICT investment that will help us estimate an initial capital stock
for an early year (see section 3.3).
Table 1: Input Output Table (IOT) and ICT asset concordance
Benchmark years IOT sectors
ISIC
industries ICT asset
1983, 1989, 1993 &
1998
Office computing machines 30 Hardware
Electronic
equipment(incl.TV) 30 Hardware
Communication equipment 32
Communication
equipment
2003 & 2007
Electronic
equipment(incl.TV) 30 Hardware
Communication equipment 32
Communication
equipment
11
As mentioned before, to extrapolate hardware investment series backward we apply the trend
in the obtained hardware series to 1999 hardware investment obtained using software/hardware
ratio from WITSA applied to software data from NAS. Alternatively, one could directly take the
series generated using commodity flow approach for the entire period (i.e. before and after 1999).
However, we opted not to do so, because, the industry concordances are not at the maximum
precision for IO tables after 2003. Hence, as discussed earlier, we constructed the hardware series
for years after 1999 using NAS based software and WITSA based software/hardware ratio. Since
the commodity flow approach produces a different estimate for hardware investment in 1999, to
keep consistency, we proportionally adjust the series based on commodity flow approach – i.e. by
applying the annual changes in commodity flow-based series to the 1999 benchmark estimate of
hardware investment.
Table 2: Comtrade and ICT asset concordance
HS revision
HS
code HS industry
ISIC
industries ICT asset
3 (1998 and
after)
714 Office machines 30 Hardware
724
Telecommunications
apparatus 32
Communication
equipment
1 (before 1998) 75 office machines, adp mach. 30 Hardware
76 Telecomm, sound equip etc. 32
Communication
equipment
Communication equipment: For the communication equipment series, we directly take the series
generated using the commodity flow approach, using data from the IO table, because the industry
description was quite consistent and clear (see Table 1). This way, we have a complete series of
ICT investment for the aggregate economy. This approach allows us to generate investment series
only for total economy, as an industry break-down is not possible with input-output table.
For all the three assets, investment series for years before 1983 (i.e. the first bench mark I-O
table), we linearly forecast the domestic availability ratios (see equation 1), until 1970. The
investment estimates obtained from this linearly interpolated series will be used to derive initial
capital stock (see section 3).
12
2.1.1 Estimates of aggregate economy ICT investment
This section presents the estimates of aggregate economy ICT investment – software, hardware
and communication – using the methodology discussed in the previous section. First we provide
the official software investment data as available from the NAS. It is evident from Figure 3 that
the nominal value of investment in software has increased significantly over the period 1999-2000
to 2012-2013. It has increased from ₹14,175 crores to ₹107,166 crores (nearly 8 times in 14 years).
However, its relative size has remained by and large stable. As a percentage of total GFCF,
software investment has declined from 2.9 percent in 1999-2000 to 2.3 percent in 2004-2005 (with
almost all years seeing a decline), and then had a steady increase until reaching 3.5 percent in
2008-2009. Since 2008, it has been fluctuating, and it hovered around 3.3 percent.
Figure 3: Software investment (₹crores) and its share in total GFCF: Aggregate Economy
Source: Authors’ computations based on National Accounts Statistics.
In Figure 4, we further provide the nominal value of total ICT investment by three asset
types – software, hardware and communication equipment – along with the share of total ICT
investment in aggregate gross value added. Note that these numbers are obtained by using the
methodology described in the previous section, except for the software investment for the period
after 1999-2000, which are official NAS series. The figure suggests a significant increase in the
total ICT investment over years. It has increased from ₹7,995 crores in 1990-1991 to ₹40,475
crores in 1999-2000 (5 times in a decade). There has been a rapid increase after that, reaching₹
2.9%
2.3%
3.5%
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
19
99
-00
20
00
-01
20
01
-02
20
02
-03
20
03
-04
20
04
-05
20
05
-06
20
06
-07
20
07
-08
20
08
-09
20
09
-10
20
10
-11
20
11
-12
20
12
-13
20
13
-14
20
14
-15
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
3.0%
3.5%
4.0%
₹C
rore
s
% o
f G
FCF
Rs. Crores Share in GFCF (%)
13
284,874 crores in 2012-2013. While the communication investment did see a remarkable increase
during 2004-2011 period, also the period of rapid expansion of mobile phones, it started declining
since then. It increased from ₹14,267 crores in 2003-2004 to ₹79,362 crores in 2010-2011, but
declined in the next two years, reaching ₹50,924 crores in 2012-2013. Hardware constitutes the
largest chunk of ICT investment in India and has also been increasing over time; hardware
investment increased from ₹23,265 crores in 2004-2005 to ₹130,723 crores in 2012-2013.
Figure 4: ICT investment (₹ crores) and its share in GFCF, Aggregate Economy
Source: Authors’ computations using data from National Accounts Statistics, WITSA and EU
KLEMS (see text).
Yet, as is the case with software investment, the relative size of ICT investment in India
i.e. ICT investment as a proportion of GFCF has not increased significantly over years. It has been
quite volatile during the decade of 1990-2010, ranging from 5.7 percent in 1990-1991 to 10.8
percent in 1994-1995. In 2000-2001, it was 9.8 percent, and then started declining continuously
reaching 6.4 percent in 2004-2005. Though it increased since then until 2008-2009, reaching its
peak at 11.6 percent, it declined since then, and continued to decline until 2012-2013, reaching 9.3
percent, close to the share it attained in 1992-1993. Thus the relative importance of ICT investment
in India has not increased significantly.
0.0%
2.0%
4.0%
6.0%
8.0%
10.0%
12.0%
14.0%
0
50,000
100,000
150,000
200,000
250,000
300,000
350,000
19
90
-91
19
91
-92
19
92
-93
19
93
-94
19
94
-95
19
95
-96
19
96
-97
19
97
-98
19
98
-99
19
99
-00
20
00
-01
20
01
-02
20
02
-03
20
03
-04
20
04
-05
20
05
-06
20
06
-07
20
07
-08
20
08
-09
20
09
-10
20
10
-11
20
11
-12
20
12
-13
20
13
-14
20
14
-15
% o
f G
FCF
₹C
rore
s
Software Communication
Hardware ICT share in GFCF
14
Figure 5: Composition of ICT investment: shares of software, hardware and
communication equipment in total ICT investment
Source: Authors’ computations using data from National Accounts Statistics, WITSA and EU
KLEMS (see text).
The distribution of ICT investment in terms of software, hardware and communication
equipment during 1990-2011, is depicted in Figure 5. Even though India’s software sector has
been dominating in its ICT exports (see Erumban and Das, 2016), the share of software in
aggregate economy ICT investment stayed at about 1/3rd on average since the 1990s. In general,
hardware investment dominated, though its share has declined over time. In 1990-1991,
communication equipment constituted about 1/4th of total ICT investment, which declined to less
than 1/5th by the end of the 1990s. However, it did see an increasing trend since the 1990s,
reaching its peak at 34 percent in 2010-2011, and since then it started declining reaching 18 percent
in the last year of the data. On average, during 2008-2012 period, both software and
communication equipment accounted for about 30% percent of total ICT investment each, whereas
hardware accounted for nearly 40 percent.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%1
99
0-9
1
19
91
-92
19
92
-93
19
93
-94
19
94
-95
19
95
-96
19
96
-97
19
97
-98
19
98
-99
19
99
-00
20
00
-01
20
01
-02
20
02
-03
20
03
-04
20
04
-05
20
05
-06
20
06
-07
20
07
-08
20
08
-09
20
09
-10
20
10
-11
20
11
-12
20
12
-13
Software
Communication
Hardware
15
2.2 Approach to estimate ICT investment in organized manufacturing
The construction of ICT series for KLEMS industries of the economy is even more challenging,
and therefore, in the current version we attempted this only for the industries in the organized (or
formal) segment of manufacturing sector. ASI schedule on fixed assets provides information on
7 individual assets. These are land; building; plant & machinery; transport equipment; computer
equipment including software; pollution control equipment and other assets. The data on computer
equipment including software is available only since 1998-1999. This has been taken as the
benchmark data for ICT investment in organized manufacturing.11 We assume it to consist of the
sum of hardware and software investment, and exclude communication equipment. It appears that
the communication equipment is still part of plant & machinery.12 We impute GFCF from the fixed
assets data using ASI tabulation procedure, i.e.
𝐺𝐹𝐶𝐹𝑖,𝑡𝑗_𝑎𝑠𝑖
= 𝑁𝑉𝐹𝑖,𝑡𝑐,𝑗_𝑎𝑠𝑖
− 𝑁𝑉𝐹𝑖,𝑡𝑜,𝑗_𝑎𝑠𝑖
+ 𝐷𝑖,𝑡𝑗_𝑎𝑠𝑖
− 𝐴𝐷𝑅𝑖,𝑡𝑗_𝑎𝑠𝑖
(2)
where GFCFi,t is the gross fixed capital formation, NVFi,tc is the net closing value of fixed assets,
NVFi,to is the net opening value of fixed assets, Di,t is the depreciation of asset i during the year,
and ADRi,t is the addition due to revaluation; all for ASI industry j, asset i(i=1: ICT
equipment)13and year t. A numeric example of this approach using the ASI schedule for food and
food manufacturing sector is provided in Appendix Table 1.
2.2.1 Estimates of ICT investment in organized manufacturing industries
Note that ASI also provides data on actual additions of an asset during the year (also see Appendix
Table 1). Instead of re-calculating GFCF using equation (2) one could directly use these series to
11 The ASI data on ICT investment has been used by previous studies, some of which are indicated earlier in this
paper. For instance, Vashisth (2017) constructs ICT capital stock using ASI firm level data on ICT investment and
Joseph and Abraham (2007) make use of the ICT investment data from ASI at 3 digit level in their regression of labor
productivity on ICT intensity. Sharma and Singh (2012), also uses investment in ICT from ASI, deflated using
machinery and equipment prices. This paper makes a comprehensive attempt to compile the ASI data on ICT and
construct measures of capital stock and capital services that are consistent with the approach followed in the India
KLEMS database. 12 Clearly, the ASI data underestimate the extent of ICT use in the organized manufacturing sector, as it only covers
the hardware and software investments (also see Vashisht, 2017). 13 A similar approach is followed to obtain GFCF in all asset types in India KLEMS, while this paper deals only with
the ICT.
16
represent investment in ICT. In figure 6, we provide the estimates of ICT investment in aggregate
manufacturing sector, obtained from ASI data using equation (2) along with the actual additions
as reported in ASI. Barring some slight deviations, by and large, these two series are quite similar
at the aggregate level. However, our preferred series is the measured GFCF series using equation
(2), which is also consistent with ASI’s own reported GFCF series for total of all assets, as we
have observed some notable differences between the two series across industries.
It is obvious from the figure that the absolute level of nominal investment in ICT in the
aggregate organized manufacturing sector has increased over time. It increased continuously from
about ₹650 crores in 1998-1999 to ₹3,122 crores in 2008-2009. After a decline in 2009-2010, it
continued to increase again reaching ₹4,144 crores in 2013-2014.
Figure 6: ICT investment in aggregate organized manufacturing sector (₹ Crores)
Note: GFCF is Gross fixed capital formation in ICT, obtained using ASI tabulation procedure,
i.e. actual additions to ICT assets – deductions during the year. Actual additions are as reported
in ASI. The numbers in this chart are consistent with ASI data, but not with national accounts.
Source: Authors’ computations using data from Annual Survey of Industries
While the absolute level of ICT investment has increased over time, it has not increased at
the same speed as the value added in the sector. The investment to value added ratio has been quite
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
1998
-99
1999
-00
2000
-01
2001
-02
2002
-03
2003
-04
2004
-05
2005
-06
200
6-0
7
200
7-0
8
2008
-09
2009
-10
2010
-11
2011
-12
2012
-13
2013
-14
GFCF ACTUAL ADDITIONS
17
volatile, and has in fact declined over years (Figure 7). On average it was 0.5 percent for the entire
period 1998-2013, with very little variation, ranging from 0.4 percent in 2000-2001 to 0.6 percent
in 2003-2004. In the last year of the data, it was only 0.42 percent of aggregate value added.
Obviously, the pace of increases in ICT use in Indian manufacturing has been relatively slower
compared to the pace of output expansion. Similar is the case even if we compare the ICT share in
total manufacturing investment, suggesting that other types of assets were growing even faster
than ICT (Figure 8).
Figure 7: ICT investment/Value added ratio: aggregate organized manufacturing sector
(%)
Note: all numbers are consistent with ASI factory sector data, and not necessarily consistent with national
accounts data.
Source: Authors’ computations using data from Annual Survey of Industries
0.3%
0.4%
0.4%
0.5%
0.5%
0.6%
0.6%
0.7%
1998
-99
1999
-00
2000
-01
2001
-02
2002
-03
2003
-04
2004
-05
2005
-06
2006
-07
200
7-0
8
2008
-09
2009
-10
2010
-11
2011
-12
2012
-13
2013
-14
18
Figure 8: Share of ICT investment in total GFCF: aggregate organized manufacturing
sector (%)
Note: all numbers are consistent with ASI factory sector data, and not necessarily consistent with national
accounts data.
Source: Authors’ computations using data from Annual Survey of Industries
Estimates of ICT investment are also made for 13 India KLEMS manufacturing sectors
(see Appendix 1 for the list of these industries and their abbreviations used in the charts). As one
would expect, more technology intensive industries such as electrical & optical equipment
manufacturing, transport equipment and machinery manufacturing industries absorb a major
chunk of ICT investment within the organized manufacturing (Figure 9). However, this
composition was quite different in the early 1990s, compared to what it is today. In 1993-1994, 22
percent of total ICT investment was in chemicals, 18 percent in textiles and 9 percent in food
products. All other industries had less than 5 percent, with wood products being the lowest and
non-metallic mineral being the highest at 5 percent. The share of textiles, however, dropped
substantially over the years, particularly until the mid-1990s, and is now 6 percent. The share of
chemicals also declined by almost 10 percentage point, reaching 13 percent in 2013-2014. The
share of machinery, transport and electrical & optical equipment, on the other hand, increased
respectively from 6 percent to 13 percent, from 7 percent to 17 percent and from 8 percent to 18
percent. Thus, together these three industries constitute almost half of the total ICT investment in
India’s organized manufacturing sector in 2013-2014.
0.5%
0.7%
0.9%
1.1%
1.3%
1.5%
1.7%
1.9%
2.1%
2.3%
2.5%19
98-9
9
1999
-00
2000
-01
2001
-02
2002
-03
2003
-04
200
4-0
5
2005
-06
2006
-07
2007
-08
2008
-09
2009
-10
2010
-11
2011
-12
2012
-13
2013
-14
19
Figure 9: Industry share in total ICT investment (%)
Note: industries are ranked in order of their relative ICT investment size in 2013-2014. All numbers are
consistent with ASI factory sector data, and not necessarily consistent with national accounts data.
Source: Authors’ computations using data from Annual Survey of Industries
3. Extending India KLEMS database to include ICT capital
In India KLEMS database, investment is used as an input to generate capital services that are
ultimately used in a growth accounting analysis to estimate input and productivity contribution to
output growth. In the case of ICT investment, it requires estimates of ICT capital stock –
calculation of which requires nominal investment and ICT prices – and its rental prices or user
cost. The approach to estimate rental prices are discussed extensively in Erumban (2008) and the
specific context of India KLEMS database is described in the India KLEMS data manual.14 In this
section, we discuss the approach to estimate ICT capital stock and rental prices very briefly, after
making the investment data consistent with National Accounts, and thus with existing India
KLEMS data.
14see https://www.rbi.org.in/Scripts/PublicationReportDetails.aspx?UrlPage=&ID=855
₹ 533 Crores
₹ 3,994 Crores
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%1
99
3-9
4
19
94
-95
19
95
-96
19
96
-97
19
97
-98
19
98
-99
19
99
-00
20
00
-01
20
01
-02
20
02
-03
20
03
-04
20
04
-05
20
05
-06
20
06
-07
20
07
-08
20
08
-09
20
09
-10
20
10
-11
20
11
-12
20
12
-13
20
13
-14
ELECTRIC&OPTIC
TRANSPORT
MACHINERY
CHEMICAL
METAL&PRODUCT
FOOD,BEV.TOBACO
TEXTILES
PAPER
NON-MET.MINERAL
RUBBER&PLAST
PETROLEUM
OTHER MFG
WOOD
ICT investment (₹ Crores)
20
3.1 Consistency with NAS
In the preceding sections, we developed a series of ICT investment for the aggregate economy and
organized manufacturing industries. As indicated earlier the aggregate economy data is fully
consistent with NAS. However, the measured ICT investment for the organized manufacturing
sector is not consistent with the aggregate economy ICT estimates for two reasons. Firstly, the
aggregate economy series is a complete series that consists of investment in software, hardware
and communication equipment, while the manufacturing series is only the sum of hardware and
software investment (distribution of these two assets are not available), and does not include
communication equipment. Secondly, all the series in the aggregate economy is benchmarked to
available data on software investment from the NAS, and is thus consistent with NAS. This is not
the case with the manufacturing series, which is solely based on ASI data. It is to be noted that
there is significant difference between the organized sector capital formation as reported in ASI
and the organized manufacturing GFCF data provided in the National Accounts. Therefore, while
developing the asset-industry investment series for non – ICT assets (such as buildings, machinery
and transport equipment), the India KLEMS has used the asset-industry distribution in the ASI
data, keeping the NAS reported GFCF data as the benchmark series.15 This way, the India KLEMS
data asset-industry investment in non/ICT assets is fully consistent with NAS data.
To ensure the NAS consistency we assume the ICT investment to total GFCF ratio in the
organized ASI manufacturing sector to the reported total GFCF data in the organized
manufacturing sector in the National Accounts. More specifically:
𝐺𝐹𝐶𝐹𝐼𝐶𝑇,𝑡𝑗
= [𝐺𝐹𝐶𝐹𝐼𝐶𝑇,𝑡𝑗_𝑎𝑠𝑖
/𝐺𝐹𝐶𝐹𝑡𝑗_𝑎𝑠𝑖
]𝐺𝐹𝐶𝐹𝑡𝑗 (3)
where𝐺𝐹𝐶𝐹𝐼𝐶𝑇,𝑡𝑗
is the gross fixed capital formation in ICT in industry j in year t, consistent
with National Accounts,𝐺𝐹𝐶𝐹𝐼𝐶𝑇,𝑡𝑗_𝑎𝑠𝑖
is the GFCF in ICT in ASI industry j, 𝐺𝐹𝐶𝐹𝑡𝑗_𝑎𝑠𝑖
is the total
GFCF (across all assets) in ASI industry j, and 𝐺𝐹𝐶𝐹𝑡𝑗 is the total GFCF across all assets in
industry j, consistent with National Accounts.The obtained total ICT investment using (3) can then
15 See the India KLEMS data manual available at: https://www.rbi.org.in/Scripts/PublicationReportDetails.
aspx?UrlPage=&ID=855
21
be split into software and hardware using the aggregate economy distribution of these two assets,
as estimated in section 2.1.1 i.e.
𝐺𝐹𝐶𝐹𝑖,𝑡𝑗
= 𝐺𝐹𝐶𝐹𝐼𝐶𝑇,𝑡𝑗
[𝐺𝐹𝐶𝐹𝑖,𝑡
𝐺𝐹𝐶𝐹𝑆𝑂𝐹𝑇,𝑡+𝐺𝐹𝐶𝐹𝐻𝐴𝑅𝐷,𝑡]] (4)
where 𝐺𝐹𝐶𝐹𝑖,𝑡𝑗
is the GFCF in asset i (software and hardware) in industry j in year t, 𝐺𝐹𝐶𝐹𝐼𝐶𝑇,𝑡𝑗
is
the GFCF in ICT (sum of software and hardware as obtained in 3) in industry j in year t and 𝐺𝐹𝐶𝐹𝑖,𝑡
is the GFCF in aggregate economy.
The NAS consistent ICT investment series generated using the above approach appears to
underestimate the total manufacturing ICT investment. As indicated in the previous section,
National Accounts provides data on investment in software by various industries, which includes
organized manufacturing as well. The total ICT investment (sum of hardware and software)
generated using equation (3) is much lower than the total software investment reported in NAS.
This is due to the differences in the way in which National Accounts and ASI compiles their GFCF
data – while the former uses an institution approach making use of enterprise level data, the latter
is based on establishment approach. This remains a major issue and needs to be tackled in the
future revisions of the data.16What is also less evident in the ASI data – which is based on
establishment approach – is how it treats ICT assets which are present at the headquarters of a firm
and not in the factory surveyed. If these assets get distributed properly among the different factories
the firms owns, the difference between establishment and enterprise/institutional approach should
be minimal, which is not the case here. A careful look at the distribution of these assets is required
to understand this re-distribution issue.
3.2 ICT prices and depreciation
Price measurement for ICT assets has been an important research topic in the literature, as the
quality of ICT capital goods has been rapidly increasing. The use of a single harmonized deflator
across countries was widely advocated and used (Timmer and van Ark, 2005; Schreyer, 2002). We
16 An alternative approach, which is not attempted in the paper yet, is outlined in Appendix 1.
22
use a harmonization procedure suggested by Schreyer (2002) where the US hedonic deflators17 are
adjusted for India’s domestic inflation rates, i.e.
∆𝑙𝑛𝑃𝐼𝑁𝐷𝐼𝐶𝑇 = ∆𝑙𝑛𝑃𝐼𝑁𝐷
𝑛−𝐼𝐶𝑇 + ∆𝑙𝑛𝑃𝑈𝑆𝐼𝐶𝑇 − ∆𝑙𝑛𝑃𝑈𝑆
𝑛−𝐼𝐶𝑇 (5)
where ∆𝑙𝑛𝑃𝑖𝐼𝐶𝑇 is the growth rate of ICT prices in country i (IND= India, US=United States) and
∆𝑙𝑛𝑃𝑖𝑛−𝐼𝐶𝑇 is the growth rate of non-ICT prices in country i. Once a NAS consistent ICT series is
developed (see the previous section), we use the harmonized ICT deflators to deflate the nominal
ICT investments and thus obtain real ICT investment series.
For the depreciation rates, following the standard practice in the literature (see Jorgenson
and Vu, 2005), a 31.5 percent for software and hardware and 11.5 percent for communication
equipment are used.
3.3 Capital stock and the contribution of ICT capital to growth
India KLEMS uses measures of capital services in its growth accounting analysis (see Das et al,
2016). This requires asset wise capital stock estimates. As is the case with all other assets, we use
a standard perpetual inventory method (PIM) to estimate ICT capital stock. Capital stock for a
given ICT asset i can be obtained using the PIM as:
𝑆𝑡𝑖 = 𝑆𝑡−1
𝑖 . (1 − 𝛿𝑖 ) + 𝐼𝑡𝑖 (6)
where 𝑺𝒕𝒊=capital stock in ICT asset i in year t, 𝜹𝒊 = geometric depreciation rate of ICT asset i and
𝑰𝒕𝒊=real investment in ICT asset i (deflated using the harmonized price deflators). This requires an
initial capital stock. Given that investing in ICT is relatively a new development, it is ideal to
assume a zero initial stock for an early year, say 1970. Due to high depreciation rates of ICT assets,
this assumption does not affect the growth rates for later years. As mentioned in section 2, we have
extended the ICT investment series for the aggregate economy until 1970, using a linearly
17
Our harmonized price deflators are based on the U.S hedonic prices, which are constructed using a hedonic regression where prices of ICT
equipment regressed on several characteristics, such as for instance processor speed, hardware size, memory etc.
23
extrapolated domestic availability ratio and the time-series of domestic availability of ICT assets,
and these series can be used to obtain an initial stock for 1970.
For the manufacturing sector, given the short time-series we have, we approximated an
initial capital stock using capital stock to investment ratio in the aggregate economy for 1992-93.
Given that the initial capital stock is estimated in a crude way, the results on growth contribution
will be reliable only after a few years, after allowing for the depreciation of a major part of first
available investment data. Fortunately, given the high rate of depreciation, the effect of initial
capital on estimated growth rate will vanish quite quickly.
The contribution of ICT capital to value added growth is obtained as the product of ICT
capital compensation share in nominal value added and the growth rate of ICT capital stock.18 The
compensation share of ICT capital is estimated by multiplying rental price of ICT capital with ICT
capital stock, with the rental price being estimated using depreciation rate, an assumed external
rate of return and investment prices, as spelled out in Das et al (2016)19.
4. Contribution of ICT to economic growth – aggregate economy and organized
manufacturing
In figure 10 we provide the contribution of ICT to value added growth in the aggregate economy
since 1996 till 2011. The whole period is divided into three sub-periods, 1996-2000, 2001-2005
and 2007-2011. On the right hand side of the figure we have the value added growth, and on the
left hand side the contribution of ICT capital to value added growth. Clearly, the value added
growth has increased over years, however, the contribution of ICT capital has not increased by a
similar magnitude. In the first period, out of more than 6 percent value added growth, only less
than 1 percentage point was due to ICT use (i.e. about 15 percent of total growth). This has declined
to 0.6 percentage points in the second period, even when value added growth increased to more
than 6.5 percent, thus lowering the relative contribution to less than 10 percent of value added
growth. In the last period, however, there has been some improvement. When the value added
18 ICT income share is obtained using ICT rental prices, which are computed using internal rate of return,
depreciation rate and ICT investment deflators (see Erumban and Das, 2016 and Erumban 2008 for detailed discussion
on the calculation of rental prices). 19 Also see the India KLEMS data manual, available at: https://www.rbi.org.in/Scripts/PublicationReportDetails.
aspx?UrlPage=&ID=855
24
growth increased to 8.3 percent, contribution of ICT increased to 1.4 percent, thus regaining its
relative contribution to about 16 percent.
A comparison of our results with that of the Conference Board Total Economy Database (TED)
suggests a similar trend in the contribution of ICT capital. According to the TED data, the relative
contributions of ICT (relative to GDP growth) has declined from 18 percent during 1996-2000 to
9 percent during 2001-2005 and then increased to just above 1/5th of the GDP growth.20
Figure 10: Contribution of ICT (hardware, software and communication equipment)
capital to value added growth, Aggregate Economy
Source: Authors’ computations using data from NAS.
20 Note that there are some differences between TED estimates and our estimates. The first is that in the TED output
and GFCF data for the aggregate are converted to calendar year (using quarterly data) and the asset distribution is
applied to those annual data to obtain asset wise data. Secondly, the price deflator for ICT used in the TED are based
on a new alternative measures developed by Byrne and Corrado (2016), while the hedonic deflators used in our paper
is based on official BEA ICT hedonics for U.S. Finally, the TED numbers are relative to ‘GDP’, whereas ours is
relative to gross value added.
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
7.0%
8.0%
9.0%
ICT contribution GDP growth
1996-2000
2001-2005
2006-2011
25
Figure 11: Contributions of ICT (hardware, software and communication equipment) and
non-ICT capital to GDP growth, Aggregate Economy
Source: The Conference Board Total Economy Database, May 2016 and authors’ computations using data
from NAS.
In Figure 11 we further compare our results for India with several other countries – both mature
and emerging market economies – using the TED data. Interestingly, India compares well with
other developing countries in regard to the contribution of ICT capital to growth. If we consider
the contributions of ICT capital and non-ICT capital to economic growth and assess the relative
importance of these two, India seem to be doing well at the aggregate level. However, compared
to mature economies, where the ICT revolution started even earlier than in most emerging market
economies, the relative importance of ICT in India and other emerging market economies is still
low.21
Given that we do not have a long series of ICT data in the database, it is appropriate to
focus on the results for most recent years, after allowing for depreciated stock in the early years.
Therefore, we provide the results on the contribution of ICT capital to value added growth in
21
In an earlier study, Papaioannou and Dimelis (2007) suggest strong ICT effects on economic growth, with the effects being larger on
developed economies than on developing economies.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Sou
th K
ore
a
Taiw
an
Un
ited
Sta
tes
Un
ited
Kin
gdo
m
Jap
an
Ger
man
y
Ch
ina
Vie
tnam
Mal
aysi
a
Ph
ilip
pin
es
Mex
ico
Sou
th A
fric
a
Ind
on
esia
Sri L
anka
Ind
ia
Ind
ia (
KLE
MS)
Non-ICT contribution ICT contribution
26
registered manufacturing only for 2001-2011 period. On average the contribution of ICT capital
(excluding telecommunication equipment) in registered manufacturing sector is about 0.2
percentage point over the period 2001-2011 (Figure 12). There has been only a marginal increase
from 2001-2005period to 2006-2011 period. Despite several industries witnessing improvement
in ICT contribution, the drop in two major industries – paper, printing and publishing industry,
and electrical and optical equipment manufacturing industries – has driven down the aggregate
contribution. These two industries had the highest ICT contribution in the first period. In the case
of paper and products the faster growth of ICT might be due to the large presence of publishing
industry (including media reproduction and newspaper), the share of which in the total value added
still relatively small, but is increasing. Electrical and optical equipment industry is an obvious
candidate to have higher ICT investment, as it is globally an ICT intensive sector. Machinery and
transport equipment sectors were the other two sectors with notable ICT contribution to value
added growth. While in both these sectors the contribution of ICT remained by and large the same,
electrical and optical equipment sector and paper and publishing sector both witnessed significant
declines in their ICT contribution.
Clearly there is significant variation across industries in ICT use. And overall there has
been some improvement – 11 out of 13 industries have witnessed an increased, albeit small in
magnitude, in the absolute contribution of ICT to growth – leading to minor increase in the
contribution of ICT to growth in organized manufacturing. Yet, the performance of registered
manufacturing, in terms of ICT use, seem to be quite negligible, when compared to the aggregate
economy, implicitly indicating that perhaps it is the services sector of the economy that has been
using the ICT more intensely. This, however, require further extension of the data to include
service sector.
27
Figure 12: Contribution of ICT (hardware and software only) capital to value added growth,
Registered Manufacturing Sector
Note: The results in this chart are made consistent with national accounts data (see text). Detailed industry
names are provided in Appendix Table 2.
Source: Authors’ computations using data from ASI and NAS.
5. Main findings
This paper made a first attempt to estimate ICT investment in Indian economy – both at aggregate
level and for individual manufacturing industries in the organized segment. Our estimates seem to
suggest an increase in the nominal value of investment in software, hardware and communication
equipment in the aggregate economy over the last 2 decades. However, as a share in GDP, it has
in fact declined. A similar pattern is observed in the case of aggregate organized manufacturing
sector as well, where we see the total ICT investment (sum of hardware and software) increases
over years, while its share in value added and total investment has been quite volatile with little
increase.
Subsequently, the paper also provided some preliminary results on the contribution of ICT
capital to value added growth in aggregate economy, and 13 India KLEMS manufacturing
industries, in the organized segment of the economy. The results suggests that while there is some
marginal increase in the ICT use and its contribution to growth in Indian manufacturing, the sector
still lags quite behind, particularly when compared to the aggregate economy. Obviously, there is
-0.2% 0.0% 0.2% 0.4% 0.6%
Contribution of ICT Capital
2001-2005
2006-2011
-10.0% -5.0% 0.0% 5.0% 10.0% 15.0% 20.0%
TRANSPORT
MACHINERY
OTHER MFG
ELECTRIC&OPTIC
WOOD
METAL&PRODUCT
PETROLEUM
PAPER
RUBBER&PLAST
NON-MET.MINERAL
CHEMICAL
FOOD,BEV.TOBACO
TEXTILES
TOTAL
Value Added Growth
28
more potential for exploiting the benefits of ICT in Indian economy, particularly in the manufacturing
sector, which can help Indian firms to improve their competitiveness.
While many factors play a role here in making Indian industries’ low ICT use, an important
one is the role of complementary investment. To embrace ICT and digital technology fully into
the production process, firms also need to invest in several complementary assets, including human
capital, training, intangible assets, and organizational restructuring. For instance, a firm level study
by Bresnahan et al. (1999) suggests that the declines in ICT price lead to increased use of a
complementary system. They observe a complementarity between skilled workers and ICT and
suggest the importance of complementary organizational investment in raising the effects of ICT
on labor demand. Such complementary investments, particularly those related to significant
internal re-organization within business organizations are often a slow process and is often a
function of culture as well. Therefore, even if the investment is made in information and
communication assets, translating it into productivity requires such complementary investment. In
the case of India, the low level of investment in ICT might itself is partly due to the low pace of
such complementary assets. Even the invested assets may not be fully utilized and translated into
growth, if not accompanied by required human capital and organizational investment – an aspect
that cannot be captured by our growth accounting approach.The results in this paper are
preliminary, and require further scrutiny and revisions. Moreover, there is still substantial room
for improving these estimates, as we are bound to make strong assumptions in the calculation
procedure.
6. Way Forward
The estimates presented in this paper may be considered as preliminary, as it involves significant
amount of assumptions and imputations. In the process of constructing these estimates, we realize
the estimates could be improved using more available information, and these will be explored in
the subsequent works. For instance, instead of using trade data from UN Comtrade (see Section
2.1), we could exploit detailed data from WITS or DGCSI. Also, given that most of the ICT
investment goods are likely to be produced in the formal/organized segment of the manufacturing
sector, instead of relying on NAS output data in constructing the domestic availability (see section
2.1), one may use detailed output data from ASI.
29
Since the commodity flow approach provides estimates of hardware and communication,
it is logical to use this series on hardware for the entire period. However, the current version, we
do not use the hardware estimates obtained from CFM for years after 1999, because of the
inaccuracy in concordance between industry classification in input-output tables and our ICT
definition. Future work may explore further possibilities of improving the commodity flow
approach and rely fully on the IO based hardware series.
Also for imputing missing data on hardware and software, we relied on hardware/software
ratio from WITSA and from the United States. An alternative option, which may be explored in
the future versions, is to rely on the firm level information on gross fixed assets in software and
hardware from CMIE’s Prowess data.
Finally, exploiting other available sources such as the Prowess and NSSO manufacturing
surveys, estimates of ICT investment may be generated for other KLEMS industries, which include
unorganized manufacturing and services industries. This would account for further heterogeneity
across industries within Indian economy. Indeed, such industry level analysis is insightful in
obtaining a broad perspective on ICT’s contribution to growth and is far better than aggregate
analysis. Yet, one should not overlook the possibility that it can still mask several minute dynamics
at the firm level. This is of particular importance when it comes to ICT’s impact in a developing
country like India, due to significant heterogeneity across firms in terms of their ICT use and
exposure – depending upon for instance their export intensity, their target consumers etc. In the
future work, one should complement the industry analysis with detailed firm-level studies. While
this paper is our first attempt to build an ICT investment series, and to quantify its contribution to
growth in Indian industries, using a standard growth accounting approach, it does not end our quest
to answer the question of ICT’s role in India’s growth. In the future, we aim to continue analyzing
the role of ICT using more comprehensive econometric tools, which would help us understand the
intensity of ICT’s effects more precisely.
30
Bibliography
Arora, A., Athreye, S., 2002.The software industry and India's economic development, Information
Economics and Policy, 14(2), 253-273.
Byrne, D and C. Corrado., 2016. ICT Asset Prices: Marshaling Evidence into New Measures, The
Conference Board Economics Program Working Paper #16-06
Baldwin, R., 2016. The Great Convergence, Harvard University Press.
Bresnahan, T.F., E. Brynjolfsson., and L. M. Hitt., 2002. Information Technology, Workplace
Organization, and the Demand for Skilled Labor:Firm-Level Evidence, The Quarterly Journal of
Economics, 117 (1), 339-376
Central Statistical Organization, The Government of India, Annual Survey of Industries, various issues
Central Statistical Organization, The Government of India, National Accounts Statistics, various issues
Central Statistical Organization., 2010. Value addition & employment generation in the ICT sector in India,
New Delhi, Ministry of Statistics and Planning, Government of India.
Commander, S., Harrison, R., Menezes-Filho, N., 2011. ICT and productivity in developing countries: New
firm-level evidence from Brazil and India, The Review of Economics and Statistics, 93 (2), 528-541.
Das, D.K., A.A. Erumban., S. Aggarwal., S. Sengupta., 2016. Productivity growth in India under different
Policy Regimes, in Jorgenson D., Timmer, M.P., Fukao, K. (Eds), The World Economy: Growth or
Stagnation?, Cambridge University Press.
De Vries, G.J., Mulder, N., Dal Borgo, M., Hofman, A. A., 2010. ICT investment in Latin America: Does
it matter for economic growth?”, in Cimoli, M., Hofman, A.A., Mulder, N. (Eds), Innovation and economic
development: The impact of information and communication technologies in Latin America, Edward Elgar,
Northampton, Massachusetts.
Erumban, A.A, and Das, D. K., 2016. Information and communication technology and economic growth
in India, Telecommunications Policy, 40 (5), 412–431
Erumban, A, A., 2008. Rental prices, rates of return, capital aggregation and productivity: Evidence from
EU and US, CESifo Economic Studies, 3 (54), 499-533.
Inklaar, R., Timmer, M.P., van Ark, B., 2008. Market services productivity across Europe and the US,
Economic Policy, 25(33), 139-194.
Jorgenson, D.W., 2001. Information technology and the US economy, American Economic Review, 91(1),
1-32.
Jorgenson, D.W., Ho, M.S., Stiroh, K.J., 2005. Information Technology and the American Growth
Resurgence, The MIT Press, Cambridge, Mass.
Jorgenson, D.W., Vu, K., 2005. Information technology and the world economy, Scandinavian Journal of
Economics, 107(4), 631–650.
Joseph, K. J., V. Abraham., 2007. Information Technology and Productivity: Evidence from India’s
Manufacturing Sector, Working Paper 389, Centre for Development Studies, Trivandrum, Kerala, India.
Kite, G., 2012. The impact of information technology outsourcing on productivity and output: New
evidence from India, International Conference On Applied Economics, 239 – 248
Kite, G., 2013. The role of information technology outsourcing on output, productivity and technical
efficiency: Evidence from Indian firms, Journal of European Economy, 12 (3), 260-285
31
Mitra, A., Sharma, C., Veganzones-Varoudakis, M., 2016. Infrastructure, ICT and firms's productivity
and efficiency: An application to the Indian Manufacturing, in F. De Beule and K. Narayanan (eds.),
Globalization of Indian Industries, India Studies in Business and Economics, Springer Science+Business
Media Singapore.Navyashree G. R., S. Bhat., 2016. ICT Investments and Growth of Small and Medium
Firms: A Study of Food Processing Industry in India, Paper presented at the Forum for Global Knowledge
Sharing, Indian Institute of Technology, Madras, http://fgks.in/images/pdf/conf/2016/Navyashree.pdf
Papaioannou, S.K., and S.P. Dimelis., 2007. Information technology as a factor of economic development:
Evidence from developed and developing countries, Economics of Innovation and New
Technology,16(3), 179–194
Saith, A., M Vijayabaskar., 2005. ICTs and Indian Economic Development: Economy, Work, Regulation,
SAGE Publications Pvt. Ltd
Schreyer, P., 2002. Computer price indices and international growth comparisons, Review of Income and
Wealth, 48(1), 15–31.
Sharma, S., N. Singh., 2012. Information Technology and Productivity in Indian Manufacturing, India
policy forum, 2012–13, 189-238
The Conference Board, 2016. The Conference Board Total Economy Database™, May 2016,
https://www.conference-board.org/data/economydatabase/
Timmer M., van Ark, B., 2005. Does information and communication technology drive EU-US
productivity growth differentials?, Oxford Economic Papers, 57(4), 693-716.
van Ark, B., Gupta, A., Erumban, A.A., 2011. Measuring the contribution of ICT to economic growth, in
van Ark, B. (Eds), The Linked World: How ICT Is Transforming Societies, Cultures, and Economies,
Ariel and FundaciónTelefónica, Madrid.
van Ark, B., Inklaar, R., McGuckin, R.H., 2003. Changing gear: Productivity, ICT and service Industries
in Europe and the United States, in Christensen, J.F., Maskell, P. (Eds.), The Industrial Dynamics of the
New Digital Economy, Edward Elgar, London.
Vashisht., 2017. Destruction or Polarization: Estimating the Impact of Technology on Jobs in Indian
Manufacturing, ICRIER working paper 334
32
Appendix 1
Another way of estimating NAS consistent ICT series for organized manufacturing is to consider
the reported NAS software data as the benchmark, and obtain hardware as a residual. That is, for
years for which the NAS software data is available (1998 onwards), consider the NAS data as the
benchmark estimates for the aggregate registered manufacturing. Hardware investment for the first
year of the data for which the software data is available in NAS can then be obtained using the
aggregate economy software/hardware ratio. This way we can obtain the total investment in ICT
for the first year as the sum of hardware and software. For all other years the total ICT investment
may be obtained by applying the growth rate of total ICT investment in ASI. The hardware series
can then be calculated as the residual after subtracting the NAS reported software investment from
this total. More formally,
𝐺𝐹𝐶𝐹𝐻𝐴𝑅𝐷,1998𝐽 =
𝐺𝐹𝐶𝐹𝑆𝑂𝐹𝑇,1998𝐽
∅1998
where 𝐺𝐹𝐶𝐹𝐻𝐴𝑅𝐷,1998𝐽
is the GFCF in hardware in 1998 in total registered manufacturing sector,
𝐺𝐹𝐶𝐹𝑆𝑂𝐹𝑇,1998𝐽
is the total registered manufacturing investment in software in 1998 and ∅1998 is
the software to hardware ratio for the aggregate economy in 1998. Subsequently, total ICT
investment in total registered manufacturing sector may be obtained as:
𝐺𝐹𝐶𝐹𝐼𝐶𝑇,1998𝐽 = 𝐺𝐹𝐶𝐹𝑆𝑂𝐹𝑇,1998
𝐽 + 𝐺𝐹𝐶𝐹𝐻𝐴𝑅𝐷,1998𝐽
And for t>1998
𝐺𝐹𝐶𝐹𝐼𝐶𝑇,𝑡𝐽 = 𝐺𝐹𝐶𝐹𝐼𝐶𝑇,𝑡−1
𝐽 [𝐺𝐹𝐶𝐹𝐼𝐶𝑇,𝑡
𝐽_𝑎𝑠𝑖
𝐺𝐹𝐶𝐹𝐼𝐶𝑇,𝑡−1𝐽_𝑎𝑠𝑖
]
Since the software data is taken as such from National Accounts, the hardware can be obtained as
a residual:
𝐺𝐹𝐶𝐹𝐻𝐴𝑅𝐷,𝑡𝐽 = 𝐺𝐹𝐶𝐹𝐼𝐶𝑇,𝑡
𝐽 − 𝐺𝐹𝐶𝐹𝑆𝑂𝐹𝑇,𝑡𝐽
While this is still a less perfect approach, the positive side of it is that it ensures complete
consistency with NAS reported software data and the trend in aggregate ICT investment in ASI.
33
Appendix Table1: ASI fixed assets schedule - food and food manufacturing (ISIC 15) in 2007-2008 (all values in Rs. Lakhs)
Types of Gross Value Depreciation Net Value
Assets
Opening
as on Addition during the
year
Deduction
&
Adjustment
during the
year
Closing
as on
Up to
year
beginning
Provid-
ed
during
the year
Up to
year end
Opening
as on
Closing
as on
---- ----
----- ------
Due to
revaluation
Actual
Addition (3+4+5-
6) (8+9) (3-8) (7-10)
1 2 3 4 5 6 7 8 9 10 11 12
1 Land 346,496 25,224 67,623 6,525 417,839 3,805 490 4,229 396,582 482,568
2 Building 1,427,116 15,543 250,910 11,518 1,651,114 399,857 81,554 459,616 1,249,114 1,424,485
3 Plant &Machinery 5,181,320 24,921 1,118,921 68,849 6,157,510 2,285,977 392,100 2,607,042 3,251,903 3,951,832
4 Transport equipment 199,423 45 59,819 13,960 228,716 101,475 30,002 116,230 155,769 177,168
5
Computer equipment
including software 63,249 9 12,781 1,099 73,871 42,548 9,146 49,867 22,594 26,005
6
Pollution Control
Equipment 54,622 2 9,697 139 64,053 25,033 3,740 28,374 31,338 37,200
7 Others 475,123 1,657 82,990 11,080 539,896 248,120 41,657 280,660 257,574 294,437
8 Sub-total (2 to 7) 7,400,853 42,177 1,535,117 106,645 8,715,160 3,103,009 558,198 3,541,789 4,968,291 5,911,128
9 Capital work in progress 474,152 0 302,488 328,053 434,456 299 215 722 498,310 439,788
10 Total (1+8+9) 8,221,501 67,401 1,905,228 441,223 9,567,455 3,107,113 558,903 3,546,740 5,863,183 6,833,484
GFCF - R + D - O + C
Note: Estimated GFCF in ICT (12 - 11 + 9 – 4) = 26,005 - 22,594 + 9,146 - 9 = 12,548, and the estimated total GFCF in food and
food manufacturing sector = 1,461,803. The reported ASI factory sector GFCF is 1,461,802.
34
Appendix Table 2: List of India KLEMS manufacturing industries
Sl. No. ISIC codes India KLEMS industries Abbreviations
1 15t16 Food , beverages and tobacco FOOD,BEV.TOBACO
2 17t19 Textiles, textile , leather and footwear TEXTILES
3 20 Wood and of wood and cork WOOD
4 21t22 Pulp, paper, paper , printing and publishing PAPER
5 23 Coke, refined petroleum and nuclear fuel PETROLEUM
6 24 Chemicals and chemical products CHEMICAL
7 25 Rbber and plastics RUBBER&PLAST
8 26 Other non-metallic mineral NON-MET.MINERAL
9 27t28 Basic metals and fabricated metal METAL&PRODUCT
10 29 Machinery, nec MACHINERY
11 30t33 Electrical and optical equipment ELECTRIC&OPTIC
12 34t35 Transport equipment TRANSPORT
13 36t37 Manufacturing nec; recycling OTHER MFG